Reflective prices a decade of solar geoengineering research at $370m if somebody coordinates it

A program officer with a grants budget and a stack of stratosphere proposals has no way to rank them. One group wants more model runs, another wants an instrument on a balloon, and nothing in the pile explains which finding unlocks the next. Reflective's road map is written for that person: a sequenced list of the experiments, studies and infrastructure it says would be needed to make informed decisions about stratospheric aerosol injection ^[1]. MIT Technology Review reported its release ^[2]. After half a century of work and at least hundreds of studies, the source says no systematic plan for closing the remaining uncertainty existed ^[15].

The number that does the persuading is the gap between the two estimates. Coordinated, the report puts the work at about a decade and around $370 million ^[3]. Uncoordinated, roughly 20 years and nearly $1.4 billion ^[4]. That is about 3.8 times the money ^[2] and roughly $1 billion of difference ^[1], and none of it buys extra knowledge. It is the price of duplicated model runs, instruments that do not share calibration, and phase-two experiments run before phase-one results arrive.

What Reflective can move itself is far smaller. Since incorporating in late 2023 it has raised more than $20 million and granted around $4 million to several dozen research groups ^[10]. The coordinated plan averages about $37 million a year ^[3], roughly nine times everything it has granted to date ^[4]. The pitch, then, is that philanthropies or government agencies should fund this, using Reflective's ordering ^[6]. The simulator and the collaborative research hub it has already built ^[11], plus the uncertainties database the road map is built on ^[12], are the assets that make that a credible claim about who holds the schedule.

The thing being pitched is better decisions, faster; the thing being done is escalating outdoor releases of sulfur dioxide, or material that converts into it, into the stratosphere, which the report argues for even as Dakota Gruener says Reflective does not advocate deployment ^[7]. Hundreds of academics have signed an open letter since 2002 calling for a ban on exactly those experiments and for an international non-use agreement, according to MIT Technology Review ^[8]. Aarti Gupta of Wageningen University, who co-initiated that effort, says the first-order questions are not technical, and that the core one is who would control a planet-altering technology and to whose ends ^[9]. A costed sequence does not answer that, and the source shows no funder, agency or decision-making body that has committed to take delivery of the outputs.

What funders actually do, when handed a sequence like this, is buy the cheap and uncontested end of it. The first phase is labelled foundational knowledge and consists of more simulation and lab work ^[14], which is fundable without touching the non-use letter, and which also cannot settle whether injected particles spread out enough to reflect sunlight or clump and drop into the troposphere ^[13].

Any line item in the plan can be sorted by whether the result would change a decision somebody is actually positioned to make, and by whether it requires putting material into the stratosphere. The box that is easy to fund and decision-irrelevant is where money drifts by default. The box that is hard to fund and decision-relevant is the one that determines whether SAI works at all, and the road map prices it without saying who signs for it.